Nov 17, 2022·edited Nov 17, 2022Liked by Dr. Patrick May
More historical and deepful database from Germany players in comparison to other Euro countries (and especially South American countries) doesn't make it unfair to the rest?
Admittedly, I dont know the decisive parameters of your algorithm, but just from my subjective point of view the calculated optimal XI seems like a rather bad joke.
Götze and Brandt as "Schienenspieler"? Müller over Musiala who arguably has been one of the best players in the world for months now? Klostermann and Kehrer as central defenders over Rüdiger?
I have no doubts there is no sympathy bias. Just always seems that German footballers are rated more highly than others from your calculations. This German squad is hardly a vintage team. Kehrer for one appears in your top XI for Germany when he is average at best for both PSG and now West Ham. I rate Germany to get to the semis but only their Bayern players seem to be truly world class. Havertz, Rudiger, Brandt, Raum etc all are a little mediocre
Well what you are pointing to is exactly our biggest strength and weakness at the same time. Goalimpact does not really care how a player is good. It just can tell that one is.
(I once even entertained the idea that Thomas Meunier is only so highly rated, because when he plays everyone else gives 10% more, because they know he is bad. Ironically that would give a high rating for him as well.)
Kehrer, Havertz and Brandt are all extremly high rated.
Raum not so much and that is why we also point to the possibility of "weak link" impacted the game more.
More historical and deepful database from Germany players in comparison to other Euro countries (and especially South American countries) doesn't make it unfair to the rest?
Good point, I'd say not necessarily unfair, but rather less precise.
Standard deviation is higher with lower number of data.
Admittedly, I dont know the decisive parameters of your algorithm, but just from my subjective point of view the calculated optimal XI seems like a rather bad joke.
Götze and Brandt as "Schienenspieler"? Müller over Musiala who arguably has been one of the best players in the world for months now? Klostermann and Kehrer as central defenders over Rüdiger?
You are right.
But the point of the Optimal XI was to integrate all best players (according to GI) into a starting XI.
The wrong formation and starting XI is our Blind Spot.
Shock horror. German based company's model rates German players higher than any other nation.
There is no sympathy bias from our side .
I have no doubts there is no sympathy bias. Just always seems that German footballers are rated more highly than others from your calculations. This German squad is hardly a vintage team. Kehrer for one appears in your top XI for Germany when he is average at best for both PSG and now West Ham. I rate Germany to get to the semis but only their Bayern players seem to be truly world class. Havertz, Rudiger, Brandt, Raum etc all are a little mediocre
Well what you are pointing to is exactly our biggest strength and weakness at the same time. Goalimpact does not really care how a player is good. It just can tell that one is.
(I once even entertained the idea that Thomas Meunier is only so highly rated, because when he plays everyone else gives 10% more, because they know he is bad. Ironically that would give a high rating for him as well.)
Kehrer, Havertz and Brandt are all extremly high rated.
Raum not so much and that is why we also point to the possibility of "weak link" impacted the game more.
A lot to learn I guess.